use "IJDRR.dta" 

Table1
Model1
reg logEM c.MPI c.Geffectivenes logpop logpop65 ,vce(robust)
Model2
reg logEM c.MPI##c.Geffectivenes logpop logpop65 ,vce(robust)
Model3
reg logEM MPI Geffectivenes logpop logpop65 logpopden loggdp lat2 long2,vce(robust)
Model4
reg logEM c.MPI##c.Geffectivenes logpop logpop65 logpopden loggdp lat2 long2,vce(robust)
Model5
reg logEM c.MPI##c.Geffectivenes logpop logpop65 logpopden loggdp logunemployment loggini loghealth lat2 long2,vce(robust)

Figure1
reg logEM c.MPI##c.Geffectivenes logpop logpop65 logpopden loggdp lat2 long2,vce(robust)
margins, at(Geffectivenes=(-2.5(0.5)2.5) MPI=(0)) vsquish
margins, at(Geffectivenes=(-2.5(0.5)2.5) MPI=(0.2)) vsquish
margins, at(Geffectivenes=(-2.5(0.5)2.5) MPI=(0.4)) vsquish
margins, at(Geffectivenes=(-2.5(0.5)2.5) MPI=(0.6)) vsquish
margins, at(Geffectivenes=(-2.5(0.5)2.5) MPI=(0.8)) vsquish
margins, at(Geffectivenes=(-2.5(0.5)2.5) MPI=(1)) vsquish
marginsplot, ylabel(6(1)14)

graph combine MPIr1.gph MPIr2.gph MPIr3.gph MPIr4.gph MPIr5.gph MPIr6.gph

Table2
Model6
reg logEM c.MPI c.icrg_qog logpop logpop65,vce(robust)
Model7
reg logEM c.MPI##c.icrg_qog logpop logpop65,vce(robust)
Model8
reg logEM c.MPI c.icrg_qog logpop logpop65 logpopden loggdp lat2 long2,vce(robust)
Model9
reg logEM c.MPI##c.icrg_qog logpop logpop65 logpopden loggdp lat2 long2,vce(robust)
Model10
reg logEM c.MPI##c.icrg_qog logpop logpop65 logpopden loggdp logunemployment loggini loghealth lat2 long2,vce(robust)

Figure2
reg logEM c.MPI##c.icrg_qog logpop logpop65 logpopden loggdp lat2 long2,vce(robust)
margins, at(icrg_qog=(0(0.1)1) MPI=(0)) vsquish
margins, at(icrg_qog=(0(0.1)1) MPI=(0.2)) vsquish
margins, at(icrg_qog=(0(0.1)1) MPI=(0.4)) vsquish
margins, at(icrg_qog=(0(0.1)1) MPI=(0.6)) vsquish
margins, at(icrg_qog=(0(0.1)1) MPI=(0.8)) vsquish
margins, at(icrg_qog=(0(0.1)1) MPI=(1)) vsquish
marginsplot, ylabel(6(1)14)

graph combine QoG1.gph QoG2.gph QoG3.gph QoG4.gph QoG5.gph QoG6.gph

Appendix B
Model1
reg logEM c.polity2 c.Geffectivenes logpop logpop65 ,vce(robust)
Model2
reg logEM c.polity2##c.Geffectivenes logpop logpop65 ,vce(robust)
Model3
reg logEM polity2 Geffectivenes loggdp logpop logpop65 logpopden lat2 long2,vce(robust)
Model4
reg logEM c.polity2##c.Geffectivenes loggdp logpop logpop65 logpopden lat2 long2,vce(robust)
Model5
reg logEM c.polity2##c.Geffectivenes logpop logpop65 logpopden loggdp logunemployment loggini loghealth lat2 long2,vce(robust)

Appendix C
reg logEM c.polity2##c.Geffectivenes loggdp logpop logpop65 logpopden lat2 long2,vce(robust)
margins, at(Geffectivenes=(-2.5(0.5)2.5) polity2=(-10)) vsquish
margins, at(Geffectivenes=(-2.5(0.5)2.5) polity2=(-6.5)) vsquish
margins, at(Geffectivenes=(-2.5(0.5)2.5) polity2=(-3)) vsquish
margins, at(Geffectivenes=(-2.5(0.5)2.5) polity2=(3)) vsquish
margins, at(Geffectivenes=(-2.5(0.5)2.5) polity2=(6.5)) vsquish
margins, at(Geffectivenes=(-2.5(0.5)2.5) polity2=(10)) vsquish
marginsplot, ylabel(6(1)14)
graph combine Polityr1.gph Polityr2.gph Polityr3.gph Polityr4.gph Polityr5.gph Polityr6.gph

Appendix D
Model7
reg logEM c.MPI##c.Stability loggdp logpop logpop65 logpopden lat2 long2,vce(robust)
Model8
reg logEM c.MPI##c.regulatory loggdp logpop logpop65 logpopden lat2 long2,vce(robust)
Model9
reg logEM c.MPI##c.ruleoflaw loggdp logpop logpop65 logpopden lat2 long2,vce(robust)
Model10
reg logEM c.MPI##c.controlofcorruption loggdp logpop logpop65 logpopden lat2 long2,vce(robust)

Appendix F
ssc install tddens
tddens Geffectivenes MPI, sgraph bgraph
Appendix G
tddens icrg_qog MPI, sgraph bgraph

